Hyperspectral image denoising and antialiasing based on tensor space and reciprocal cell

Zhang Aiwu; Zhao Jianghua; Zhao Ningning; Kang Xiaoyan; Guo Chaofan

doi:10.3788/IRLA201847.1026002

Conventtrial denoising and antialiasing algorithms are usually for single band images. Previously, numerous studies have only designed for single band images. Aiming at the data characteristics of hyperspectral image and the influence of noise and aliasing on the image, a multidimensional filtering algorithm combining tensor and reciprocating cells was proposed and applied to denoising and antialiasing of hyperspectral images. The method introduced the tensor, and the hyperspectral image data was regarded as the third-order tensor expression. The reciprocal cell was used to obtain the spectrum extrapolation which containd less image aliasing and noise. From the point of view of the minimum mean square error, the algorithm alternately iterated to solve the three directions of the filter, and finally completed the image filtering. The algorithm could effectively reduce the image aliasing and noise under the premise of ensuring the consistency of image space and spectral information. The effectiveness of the proposed algorithm was proved by comparing with multiple sets of hyperspectral data of the two-dimensional Wiener filter algorithm and tensor multidimensional denoising algorithm.

Hyperspectral image denoising and antialiasing based on tensor space and reciprocal cell

1. Key Laboratory of 3D Information Acquisition and Application of Ministry of Education,Capital Normal University,Beijing 100048,China;

2. Engineering Research Center of Spatial Information Technology,Ministry of Education,Beijing 100048,China

Received Date: 2018-05-07

Rev Recd Date: 2018-06-12

Publish Date: 2018-10-25

Key words:

Abstract: Conventtrial denoising and antialiasing algorithms are usually for single band images. Previously, numerous studies have only designed for single band images. Aiming at the data characteristics of hyperspectral image and the influence of noise and aliasing on the image, a multidimensional filtering algorithm combining tensor and reciprocating cells was proposed and applied to denoising and antialiasing of hyperspectral images. The method introduced the tensor, and the hyperspectral image data was regarded as the third-order tensor expression. The reciprocal cell was used to obtain the spectrum extrapolation which containd less image aliasing and noise. From the point of view of the minimum mean square error, the algorithm alternately iterated to solve the three directions of the filter, and finally completed the image filtering. The algorithm could effectively reduce the image aliasing and noise under the premise of ensuring the consistency of image space and spectral information. The effectiveness of the proposed algorithm was proved by comparing with multiple sets of hyperspectral data of the two-dimensional Wiener filter algorithm and tensor multidimensional denoising algorithm.

HTML

Reference (19)

[1]	Chang C, Du Q. Estimation of number of spectrally distinct signal sources in hyperspectral imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2004, 42(3):608-619.
[2]	Xu Ping, Xiao Chong, Zhang Jingcheng, et al. Determination of plant hyperspectral data based on grouped three-dimensional discrete cosine transform dictionary[J]. Acta Optica Sinica, 2017(6):0630003. (in Chinese)
[3]	Xu Dong, Sun Lei, Luo Jianshu. Hetero-spectral hyperspectral remote sensing image denoising combined with NAPCA and complex wavelet transform[J]. Infrared and Laser Engineering, 2015, 44(1):327-334.
[4]	Huang Wei, Gao Taichang, Liu Lei, et al. A method for noise reduction of hyperspectral data based on improved cumulative variance[J]. Spectroscopy and Spectral Analysis, 2016, 36(11):3625-3629. (in Chinese)
[5]	Yu Zhenmiao, Yang Ming. High spectral image denoising algorithm combining spatial similarity and RPCA[J]. Journal of Nanjing University (Natural Science), 2017, 53(3):518-524. (in Chinese)
[6]	Zhang H, He W, Zhang L, et al. Hyperspectral image restoration using low-rank matrix recovery[J]. IEEE Transactions on Geoscience Remote Sensing, 2014, 52(8):4729-4743.
[7]	Almansa A, Durand S, Roug B. Measuring and improving image resolution by adaptation of the reciprocal cell[J]. Journal of Mathematical Imaging and Vision, 2004, 21(3):235-279.
[8]	Zhang Zhi, Xia Deshen. Image restoration of reciprocal cell combining with complex wavelet[J]. Acta Electronica Sinica, 2008, 36(10):1979-1985. (in Chinese)
[9]	Zhang Yuhui, Tang Yang, Chen Qiang, et al. Improve the effective resolution of oblique-mode remote sensing image[J]. Journal of Computer-Aided Design and Graphics, 2009, 21(2):243-249. (in Chinese)
[10]	Zhang Aiwu, Du Nan, Kang Xiaoyan, et al. Non-linear transformation and information-related hyperspectral adaptive band selection[J]. Infrared and Laser Engineering, 2017, 46(5):0538001. (in Chinese)
[11]	Wu Z, Wang Q, Jin J, et al. Structure tensor total variation-regularized weighted nuclear norm minimization for hyperspectral image mixed denoising[J]. Signal Processing, 2017, 131(1):202-219.
[12]	Yang Xinfeng, Hu Xunuo, Nian Yongjian. High-spectral image compression algorithm based on classification[J]. Infrared and Laser Engineering, 2016, 45(2):0228003.
[13]	Tang Zhongqi, Fu Guangyuan, Chen Jin, et al. Sparse representation and classification of hyperspectral images based on multi-scale segmentation[J]. Optical Precision Engineering, 2015, 23(9):2708-2714. (in Chinese)
[14]	Chen Z, Karim M A, Hayat M M. Elimination of higher order aliasings by multiple interlaced sampling[J]. Optical Engineering, 1999, 38(38):879.
[15]	He Yanfei. Remote sensing image restoration based on sparse representation and adaptive inverse cell[D]. Nanjing:Nanjing University of Information Science and Technology, 2015.
[16]	Zhang Zhi, Xia Deshen. Image resolution based on hybrid reciprocal cell-wavelet[J]. Journal of Computer-Aided Design and Computer Graphics, 2008, 20(4):512-519. (in Chinese)
[17]	Damien Muti, Salah Bourennane. Multidimensional filtering based on a tensor approach[J]. Signal Process, 2005, 85(12):2338-2353.
[18]	Damien Muti, Salah Bourennane. Noise removal from hyperspectral images by multidimensional filtering[J]. IEEE Geoscience and Remote Sensing Letters, 2008, 46(7):2061-2069.
[19]	Wang Zhongmei, Yang Xiaomei, Gu Xingfa. High-spectral image denoising algorithm of tensor groups[J]. Acta Geodaetica et Cartographica Sinica, 2017, 46(5):614-622. (in Chinese)

Hyperspectral image denoising and antialiasing based on tensor space and reciprocal cell

doi: 10.3788/IRLA201847.1026002

Abstract

References

Proportional views

通讯作者: 陈斌, bchen63@163.com

Article Metrics

Related

Proportional views